Mobile Multimedia Services in the Cloud

Mobile Multimedia Services in the Cloud Von der Fakultät für Mathematik, Informatik und Naturwissenschaften der Rheinisch-Westfälischen Technischen Hochschule Aachen zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften genehmigte Dissertation vorgelegt von Diplom-Ingenieur Dejan Kovachev aus Strumica, Mazedonien Berichter: Universitätsprofessor Dr. rer. pol. Matthias Jarke Universitätsprofessor Dr. Ing. Juan Quemada Vives PD Dr. rer. nat. Ralf Klamma Tag der mündlichen Prüfung: 14. Mai 2014 Diese Dissertation ist auf den Internetseiten der Hochschulbibliothek online verfügbar. Abstract Cloud computing is a paradigm shift promising a utility-based delivery of storage and pro- cessing capacity, services, and software over the Internet. In essence it aims to reduce costs, facilitate self-automated systems and decouple service delivery from underlying technology. Thus, the cloud paradigm empowers customers with the ability to focus on creating novel services alleviating the burden of software and hardware resource provisioning. The success of cloud computing in the domain of enterprise applications has sparked increasing interest in applying the same principles to the provision of mobile multimedia services. However, the potential benefits are far from being achieved, despite the rapid growth in popularity and omnipresence of mobile multimedia applications. The combination of cloud computing and mobile multimedia is non-trivial, and many aspects from system, mobile multimedia and user perspectives need to be considered. For example, mobile applications in the cloud involve a trade-off in terms of what should run on the device and what in the cloud, which is contingent to the application type, the device capability, data locality and the operating environment (network bandwidth, delay, cloud availability). Moreover, the traditional server/client programming models fail to provide seamless cloud execution in volatile mobile networks. Furthermore, distant cloud data centers induce prohibitive latency for certain classes of interactive mobile applications such as 3D games and augmented reality. This dissertation investigates ways to efficiently apply the concepts of the emerging cloud computing paradigm in the design, development and delivery of mobile multimedia services. It describes an information systems architecture called CAELUS (Cloud Architecture for Enabling Mobile Multimedia Services) which includes both conceptual models and a concrete software platform. The conceptual models capture specific requirements for efficient building of mobile multimedia cloud services and guide the creation of the software platform (i5Cloud) which serves as a test bed of the CAELUS architecture. The contributions of this dissertation, in a addition to a comprehensive survey of literature, comprise a design view, platform and abstraction levels that lower the barrier for mobile multimedia services to leverage the clouds. Several case studies have evaluated the CAELUS-based development and delivery of mobile multimedia cloud services. In particular, the case studies were conveyed in the application domains of technology-enhanced learning, digital documentation in cultural heritage and human-computer interaction. Prototype applications together with technical evaluations and user studies demonstrate the validity and applicability of the architecture and the conceptual approach. iii Kurzfassung Cloud Computing ist ein Paradigmenwechsel zur Bereitstellung von Speicher- und Ver- arbeitungskapazitäten, Diensten und Software über das Internet. Im Wesentlichen wird dadurch eine Kostenreduzierung, sowie die Entkopplung von Servicebereitstellung und zugrundeliegender Technologie erreicht. Somit befähigt Cloud Computing Kunden dazu, sich vollkommen auf die Entwicklung neuartiger Dienste zu konzentrieren, ohne sich um die Bereitstellung benötigter Software- und Hardware-Ressourcen kümmern zu müssen. Der Erfolg des Cloud Computing in Unternehmensanwendungen hat ein gesteigertes Interesse daran ausgelöst, die gleichen Prinzipien bei der Bereitstellung von mobile Multimedia- Diensten anzuwenden. Jedoch konnten die erwarteten Vorteile trotz des schnellen Wach- stums an Popularität und Allgegenwärtigkeit von Multimedia-Applikationen bisher noch lange nicht ausgeschöpft werden. Die Kombination von Cloud Computing und Multimedia-Diensten ist nicht trivial und erfordert die Betrachtung vieler Aspekte aus den Perspektiven des Gesamtsystems, des Nutzers und mobiler Medien. Beispielsweise wägen mobile Cloudanwendungen ab, was auf dem Gerät und was in der Cloud ausgeführt werden soll. Diese Entscheidung hängt von der Art der Applikation, der Geräteleistung, der Datenlokalität und der Betriebsumgebung (Netzwerk-Bandbreite, Verzögerung, Cloud Verfügbarkeit) ab. Darüber hinaus scheitert das traditionelle Client/Servermodell an nahtloser Cloud-Ausführung in unbeständigen Mobilfunknetzen. Außerdem verursachen entfernte Cloud-Rechenzentren Latenzen, die für bestimmte Klassen von interaktiven mobilen Anwendungen wie 3D-Spiele und Augmented Reality impraktikabel sind. Die Dissertation eruiert Methoden zur effizienten Anwendung von Konzepten des aufstreben- den Cloud Computing Paradigmas auf Design, Entwicklung und Bereitstellung mobiler Multimedia-Dienste. Sie beschreibt eine Informationssystemarchitektur namens CAELUS (Cloud Architecture for Enabling Mobile Multimedia Services), die sowohl konzeptuelle Modelle als auch eine konkrete Softwareplattform in sich vereint. Die konzeptuellen Mod- elle erfassen spezifische Anforderungen für den effizienten Aufbau mobiler multimedialer Cloud-Dienste und dienen als Richtlinien zur Erstellung der Softwareplattform (i5Cloud), die das Testbett für die CAELUS Architektur darstellt. Die Beiträge dieser Dissertation beinhalten neben einem umfassenden Literaturüberblick eine Designsicht, eine Plattform sowie geeignete Abstraktionsschichten, die durch mobile Multimedia-Dienste Hürden auf dem Weg zur vorteilhaften Nutzung des Cloud Computing überwinden. Die CAELUS-basierte Entwicklung und Bereitstellung mobiler Multimedia-Dienste wurde in mehreren Fallstudien evaluiert. Besonders wurden die Fallstudien in den Anwendungs- domänen von technologiegestütztem Lernen, digitale Dokumentation des kulturellen Erbes v und der Mensch-Computer-Interaktion durchgeführt. Prototypische Anwendungen inklu- sive begleitender technischer Evaluierungen und Nutzerstudien belegen die Validität und Anwendbarkeit der Architektur und ihres konzeptuellen Ansatzes. vi Acknowledgments I am most grateful towards each and every person who has taken an interest in this work, be it brief or lasting throughout the process. First of all, I wish to express my deep gratitude to Prof. Matthias Jarke for his guidance and advice during my years at the Chair of Information Systems and Databases, RWTH Aachen University. His experience and insight into research and academics have supported me in delivering this dissertation. I was lucky to enjoy the affiliations with the Chair, B-IT Research School and the UMIC excellence cluster. At the same time, I am highly indebted to Dr. Ralf Klamma for recommending me to pursue an academic career and for giving me the unique opportunity to work within his team. He constantly advised, encouraged and inspired my research and most importantly became a close friend. Our philosophical discussions often extended my view on the world. I feel extremely lucky to have worked under his guidance and I am thankful for all the confidence he had in me, and for the freedom, responsibilities and unselfish generosity I received. In addition, I like to thank Prof. Quemada-Vives from Spain for serving on the dissertation committee. I am particularly grateful to my math teacher Gjorgji Serafimov. What my elementary teacher Mimoza Serafimova with arduous efforts initiated in me regarding learning, Gjorgji Serafimov pick it up and molded it into a desire and skills for the beautiful world of mathematics – things that have delineated my life. Over the last years, it was a pleasure to have colleagues around who made life at the office more fun and with whom I worked together in one way or the other – thanks to Georgius Toubekis, Istvan Koren, Dr. Michael Derntl, Dr. Milos Kravcik, Dr. Khaled Rashed, Zinayida Petrushyna, Anna Hannemann et al. Besides the above, special thanks are to my colleagues who have also been close friends. The fruitful research collaboration with Dr. Yiwei Cao extended to a wonderful friendship. The ad-hoc scientific chats with Dr. Manh Cuong Pham in the halls of the Chair, slowly grew in a joyful companionship in sports, partying and research. Many thanks are to my “rommie” Dominik Renzel, who I have shared office with and with whom it was always pleasure to collaborate. The last years at the Chair have been filled with even more good moments and laughter thanks to Petru Nicolaescu. He proved to care about my dissertation by being critical while proof-reading this document and listening to my many final talk rehearsals. Many thanks also to Daniele Gloeckner, Claudia Puhl and Gabriele Hoeppermanns for helping me cut through all the red tape that comes with the workplace. vii I have been fortunate to always have students around with a genuine interest in mobile cloud computing research. In particular, I appreciate the collaboration with Gökhan Aksakali, Tian Yu, Ke Li, Roman Brandt, and Ghislain Manib Mbogos. Many thanks to Reinhard Linde for all his help in solving hardware and software issues of my projects. I would like to thank Tatjana

Load more